The primary specific aim of this proposal is to investigate the mechanisms by which the carcinogenic mutagen aflatoxin B-1 (AFB1) induces base substitution and frameshift mutations at a high efficiency. Our results to date have revealed that the primary DNA sequence as well as the helical structure of DNA influence the reaction specificity of AFB1. We have also described the sequence-specific effects of AFB1 adducts on template function in vitro, and on frameshift mutagenesis in vivo. These results, in concert with current concepts on bulky adduct induced base substitutions and the Streisinger model for spontaneous frameshifts have led us to propose a general model for AFB1 mutagenesis which argues that many of the base substitution and frameshift mutations induced by AFB1 arise as alternative expressions of the sequence-specific events associated with the bypass of AFB1 lesions in vivo. According to this model, replication blocks at AFB1 damage sites are overcome by either a loop-out of an adducted guanine, or by a base incorporation opposite the lesion; however, the lesion promotes, at a high frequency, a second mutation which appears to be induced by strand slippage at sequences close to the lesion. This hypothesis satisfactorily accounts for the striking features of AFB1 induced frameshift mutations which include a high proportion (60%) of A:T bp 'additions' within G:C bp runs and a high proportion (up to 35%) of complex frameshifts where the frameshift is accompanied by a vicinal base substitution. Our model can also account for the effects of RecA, SOS and MucAB activities on both the frequency and specificity of mutations, and offers an explanation for the nearly equal efficiency with which AFB1 reverts Salmonella TA100 and TA98 tester strains. The model links current concepts on base substitution mutagenesis by bulky lesions to the Streisinger hypothesis for spontaneous frameshifts. An attractive feature of this model is that it makes a number of specific predictions testable by presently available technology. These predictions will be tested by in vitro modification of derivatives of M13mp8 DNA by activated AFB1 by methods that include site- and domain-specific introduction of adducts followed by transfection into E. coli hosts. Mutants will be identified by their lacZ phenotype, resistance to restriction or hybridization specificity. In vitro tests for sequence-specific predicted slippage events will also be carried out. The level of resolution of all the above approaches is sufficient to permit the development of alternative models, in the event our predictions were unfulfilled.